In cases where there are more clusters a loop with more iterations should be performed.

Hi Luke, 

thanks for reaching out! I don't think there is a general rule for how many iterations should be performed. In this case, we are drawing from a marketing knowledge and the number of data points. We are also using the knowledge from the hierarchical clustering. In this case, based on the number of people and the number of features we have, having more than 10 clusters of clients is unlikely.

When you have new data, I'd recommend trying out a loop with more iterations just in case and analyzing the results. K-means works relatively fast, so you'd be able to perform such a check without much computational expense. 

Best, 

365 Eli